Comparison of prognostic accuracy of HCC staging systems in patients undergoing TACE

Abstract

Purpose

To compare the prognostic power of commonly used staging systems of hepatocellular carcinoma (HCC) for predicting overall survival after transarterial chemoembolization (TACE).

Materials and methods

This retrospective single center study included patients with HCC who underwent TACE between 2008 and 2019 in a single tertiary care center. After initial screening of 408 consecutive patients, 317 patients with HCC treated with conventional or drug-eluting beads-TACE were included. Five HCC staging systems (Barcelona Clinic Liver Cancer, Hong Kong Liver Cancer, Japan Integrated Staging, Cancer of the Liver Italian Program and Okuda) were compared using Kaplan Meier survival analysis and a log-rank test with overall survival (OS) as the study endpoint. Uni- and multivariate analyses of system-specific variables were applied to stratify outcomes and compare the ability to predict OS of patients after TACE. Four different measures were used to assess the homogeneity (Likelihood ratio:LR), discriminatory ability (linear trend:LT and C-index) and explanatory ability (Akaike Information Criterion:AIC).

Results

The OS of the total cohort was 29.8 months. In terms of prognostic stratification, the BCLC staging system had the best performance (LT: 8.209, LR: 26.639, AIC: 317, c-index: 0.818) compared to HKLC (LT: 10.919, LR: 25.802, AIC: 443, c-index: 0.835), JIS (LT: 4.611, LR: 16.880, AIC: 449, c-index: 0.548), CLIP (LT: 6.738, LR: 13.109, AIC: 501, c-index: 0.782), and Okuda (LT: 7.185, LR: 0.760. LR: 16.356, AIC: 487, c-index: 0.760).

Conclusion

Across five commonly utilized international staging systems, the BCLC staging system demonstrated the greatest prognostic accuracy with respect to predicting OS of patients undergoing TACE.